Plasma Heating Systems

In a fusion reactor, heating the plasma above 100 million degrees requires combining several heating devices with different characteristics. This page explains, from a device-engineering standpoint, what parts each heating system is made of, where the engineering difficulties lie, and what performance real machines actually achieve. The physical principles of why heating works are covered in Principles of Plasma Heating; here we focus on the hardware.

Start with Intuition (High School)

When you boil water in a kettle, the simplest method is to run electricity through a heating element so it gets hot. A tokamak has a similar mechanism: driving a large current through the plasma itself generates heat through electrical resistance. This is called ohmic heating. The trouble is that the hotter a plasma gets, the more easily it conducts electricity (the lower its resistance). Once the resistance disappears, so does the heating, so this method alone can only reach about one third of the temperature needed for fusion.

That is why we need additional heating devices that inject energy from the outside. There are three main approaches. The first heats the plasma by firing fast particles into it, like shooting billiard balls (NBI). The second heats it by applying radio waves, like a microwave oven, letting electrons absorb the waves (ECRH). The third also uses radio waves, but at the lower frequency matched to heavy ions (ICRF).

All three share the challenge of reliably delivering energy deep into the plasma, and each has its own strengths and clever tricks. Below we look in turn at what parts make up these three devices.

Understand the Physics (Undergraduate)

The makeup of neutral beam injection (NBI)

Neutral beam injection (NBI) is a device that fires large numbers of high-energy neutral hydrogen atoms into the plasma. The key is using neutral, that is uncharged, atoms: charged particles would be bent by the magnetic field and bounced back at the plasma surface, but neutral atoms can ignore the field lines and fly straight into the core. After they arrive, they collide with particles in the plasma, become ionized, get trapped by the magnetic field, and share their energy.

The device is built as a row of parts arranged in a line. First an ion source produces large quantities of hydrogen ions, and an accelerator applies tens to hundreds of kV to accelerate them in one go. Up to this point the particles are charged, so an electric field can accelerate them. Next they pass through a neutralizer, a chamber filled with a thin hydrogen gas, where the fast ions exchange electrons with the gas molecules and turn into neutral atoms. Some ions fail to neutralize, so a residual ion deflection magnet applies a magnetic field to bend them away and dump them onto a cooled metal plate called a beam dump. Only the atoms that were successfully neutralized pass through the drift tube and reach the plasma.

The larger the beam energy $E$, the deeper it penetrates the plasma, but here lies the limit of positive-ion NBI. When neutralizing positive ions (protons or hydrogen molecular ions), the neutralization efficiency drops sharply as the energy rises, falling to roughly 30 % or less above $100 \ \mathrm{keV}$. Most of the accelerated energy ends up being thrown away in the beam dump, which worsens the overall efficiency of the device.

Why negative-ion NBI is needed

For this reason, large devices use negative ions (hydrogen ions carrying one extra electron, $\mathrm{H}^-$ or $\mathrm{D}^-$) instead of positive ions. Because the binding of the extra electron in a negative ion is weak, the neutralization efficiency stays around 60 % even at high energies. ITER requires energies as high as $1 \ \mathrm{MeV}$, and in this range there is no realistic alternative to negative-ion NBI.

Producing negative ions efficiently is the key engineering difficulty. In a negative ion source, a thin layer of cesium is deposited on the electrode surface to lower its work function (the energy needed to pull out an electron), so that hydrogen atoms are more likely to pick up an electron at the surface and become negative ions. However, the extra electron is easily stripped off, so managing the unwanted electrons and particles that remain along the beam path is difficult, and delivering a large current stably over long durations is a challenge.

Electron cyclotron resonance heating (ECRH) and the gyrotron

Electron cyclotron resonance heating (ECRH) applies millimeter waves ($100$ to $200 \ \mathrm{GHz}$) matched to the rotation frequency (cyclotron frequency) at which electrons circle the magnetic field, so that electrons absorb energy resonantly. It is like a giant microwave oven, but the frequency is dozens of times higher than a household microwave ($2.45 \ \mathrm{GHz}$).

The device that generates these millimeter waves is a vacuum tube called a gyrotron. An electron beam emitted from an electron gun is made to rotate inside a strong magnetic field, and inside a cavity the energy of the electrons’ rotational motion is converted into electromagnetic waves. The exit window is made of CVD diamond (synthetic diamond made by chemical vapor deposition). Diamond conducts heat well and absorbs little of the millimeter wave, so it can withstand continuous operation at the $1 \ \mathrm{MW}$ class. Developing this window material was the key that made high-power continuous operation possible. Energy-recovery gyrotrons (which recover the remaining energy at a collector) achieve device efficiencies above 50 %.

The millimeter waves produced by the gyrotron are carried to the plasma by a transmission line. Because millimeter waves have a short wavelength (about $1.8 \ \mathrm{mm}$ at $170 \ \mathrm{GHz}$), the system uses waveguides that transmit them with low loss by reflecting off mirrors inside metal tubes, along with arrays of focusing mirrors. Thanks to the short wavelength, the heating can be localized to just the targeted spot, which also makes it useful for control that suppresses the magnetic-island instability known as the neoclassical tearing mode (NTM). Another advantage is that changing the mirror angle moves the heating location.

Ion cyclotron range of frequency heating (ICRF) and the antenna

Ion cyclotron range of frequency (ICRF) heating uses electromagnetic waves in the rotation frequency band of heavy ions ($20$ to $100 \ \mathrm{MHz}$, close to the FM radio and shortwave broadcast bands) to shake the ions directly. Amplifier circuits using tetrodes (four-electrode vacuum tubes) have a proven track record as the high-frequency source, and the conversion efficiency of around 60 to 70 % makes ICRF one of the more power-efficient heating devices.

What is difficult, on the other hand, is the antenna coupling (how power is loaded onto the plasma). The antenna is placed along the wall on the low-field (outer) side of the torus, and it consists of loop-shaped radiating straps and a Faraday shield that blocks unwanted electrostatic fields. To load the waves onto the plasma, the distance between the antenna and the plasma surface must be kept small, but the density at the plasma surface (edge region) changes moment to moment with the operating conditions. If the distance or density shifts, power bounces back toward the antenna, lowering the coupled efficiency and generating impurities around the antenna from the reflected power. Matching the impedance while tracking these changing boundary conditions is the central engineering challenge of ICRF.

Deepen the Theory (Graduate)

Pinning down each method’s performance with numbers reveals where it fits in the overall system design. The demonstrated values and roles that serve as rough guides are as follows.

Method	Frequency / energy band	Guide to power / conversion efficiency	Main uses
NBI (negative ion)	$\sim 1 \ \mathrm{MeV}$	Wall-plug efficiency 25 to 30 %	Main heating, rotation drive, current drive
ECRH	$100$ to $200 \ \mathrm{GHz}$	Gyrotron efficiency around 50 %	Local heating, NTM control, start-up assist
ICRF	$20$ to $100 \ \mathrm{MHz}$	Conversion efficiency 60 to 70 %	Ion heating, minority heating

Here wall-plug efficiency refers to the fraction of the power drawn from the outlet that actually reaches the plasma. For the $1 \ \mathrm{MeV}$ negative-ion NBI system this value is roughly 25 to 30 %. Because part of the energy used for acceleration is lost in neutralization and beam transport, it stays at this level. For ECRH the conversion from the power supply to millimeter waves (the gyrotron efficiency) is around 50 %, but transmission-line losses add on top, so the wall-plug efficiency varies from device to device. ICRF has a high power-supply efficiency, but once the coupling loss from the antenna to the plasma is included, the effective efficiency depends on the operating conditions.

When these are integrated into a single machine, the key to the design is that each device excels at a different job. Because NBI carries in the particles themselves, it can drive plasma rotation and current in addition to heating, making it the mainstay of non-inductive current drive. Because ECRH can move its heating location freely with mirrors, it suits local control of MHD instabilities and start-up assist in the early, low-density phase of operation. ICRF can selectively heat a particular ion species through a technique called minority heating, and it is also used in experiments that mimic the behavior of $\alpha$ particles. Real machines combine these, shifting the allocation phase by phase to shape both the heating and the current profile.

ITER’s heating system configuration

ITER is equipped with all three methods. The configuration comprises negative-ion NBI at $33 \ \mathrm{MW}$ ($1 \ \mathrm{MeV}$, long pulses up to $3600 \ \mathrm{s}$), ECRH at $170 \ \mathrm{GHz}$ delivering $20 \ \mathrm{MW}$, and ICRF at $40$ to $55 \ \mathrm{MHz}$ delivering $20 \ \mathrm{MW}$, planning for additional heating power exceeding $70 \ \mathrm{MW}$ in total. To achieve this scale and long-pulse operation, elements such as the $1 \ \mathrm{MeV}$ negative ion source for NBI, continuous operation of the gyrotron, and higher voltage tolerance for the ICRF antenna have each been developed at full machine scale. JT-60SA also relies on NBI and ECRH as its mainstays, taking on the demonstration of long-pulse operation and heating scenarios.

Research Frontier (PhD)

In today’s heating-system engineering, achieving both high power and long pulses is a shared goal. To make fusion power generation viable, the heating devices must run continuously for minutes to steady state, and managing the heat load on the components is the biggest challenge. The main topics under research are as follows.

Negative ion source development is the heart of NBI. The challenge is the technology to extract a high-current negative ion beam at the $1 \ \mathrm{MeV}$ class stably, uniformly, and over long durations, and research covers control of cesium vapor supply and recovery, suppression of the unwanted electrons generated inside the beam (co-extracted electrons), and the reliability of high-voltage insulation.

For gyrotrons, work is advancing toward even higher frequencies and higher power. Raising the frequency widens the range of magnetic field and density that can be heated, and research themes include tunable-frequency gyrotrons that can switch among multiple frequencies, single-tube output exceeding $1 \ \mathrm{MW}$, and reducing losses across the whole transmission line. For ICRF, continuing efforts address antennas that maintain coupling by tracking the fluctuating plasma surface (ITER-type designs and real-time matching) and suppressing impurity generation caused by reflected power.

Keywords you often encounter in the literature include wall-plug efficiency, current drive, minority heating, neoclassical tearing mode control (NTM control), and the negative ion source. These device technologies are increasingly designed and evaluated not just for their standalone performance but as part of operating scenarios that integrate heating, current drive, and instability control.

Check Your Understanding

Q1. Why can ohmic heating alone not reach the temperature required for fusion?

A Because a plasma's electrical resistance grows as its temperature rises, causing excessive heating that destroys the device B Because a plasma's electrical resistance falls as its temperature rises, so resistive heating stops working C Because increasing the plasma current disturbs the magnetic field and breaks confinement D Because ohmic heating warms only the electrons and cannot heat the ions at all

Q2. Which is the correct reason for using neutral atoms in NBI?

A Because neutral atoms are easy to accelerate with a magnetic field and can be raised to high energy in one go B Because neutral atoms are uncharged, so they are not bent by the magnetic field and reach straight into the plasma core C Because neutral atoms reflect off the plasma surface and can heat only the edge region D Because neutral atoms are lighter than ions and reach deep with less energy

Q3. Why does negative-ion NBI become necessary in large devices?

A Because the neutralization efficiency of positive ions drops sharply at high energy, while negative ions keep their neutralization efficiency even at high energy B Because negative ions are easier to produce than positive ions, making the device cheaper C Because positive ions are not bent by the magnetic field and only negative ions can be controlled by it D Because negative ions can be delivered to the plasma as they are, without neutralization

Q4. Why is CVD diamond used for the gyrotron window material?

A Because diamond does not conduct electricity and can reflect and focus millimeter waves onto the plasma B Because diamond is cheap and can be mass-produced, reducing replacement frequency C Because diamond has high thermal conductivity and absorbs little of the millimeter wave, so it withstands the heat generated in 1 MW class continuous operation D Because diamond is strong against magnetic fields and does not deform even in strong fields

Q5. Which is the correct central engineering challenge of ICRF?

A Its high-frequency source has lower conversion efficiency than other methods, making a larger power supply unavoidable B The millimeter-wave wavelength is too short to reach the targeted spot C The antenna must be placed on the high-field side, and there is not enough installation space D Maintaining the coupling (impedance matching) between the antenna and the plasma while tracking the fluctuating density of the plasma edge

Related Topics

• Principles of Plasma Heating - The physical principles of why each heating method works
• ITER Project - The world’s largest tokamak experimental reactor, integrating NBI, ECRH, and ICRF
• JT-60SA Project - Demonstration of long-pulse heating scenarios using NBI and ECRH